![\"Open](\"https://colab.research.google.com/assets/colab-badge.svg\"){kind=icon}
"
]
},
{
"cell_type": "markdown",
"source": [
"# Lab 10 - Vanilla CNN and Fine-Tune VGG16 - for Dogs and Cats Classification"
],
"metadata": {
"id": "OoeWP2OY6z_v"
}
},
{
"cell_type": "markdown",
"source": [
"### Goal : Take an existing model that performs a similar task of image classification to achieve and fine-tuning it for the specific task at hand which is of classifying dogs and cats."
],
"metadata": {
"id": "b2dka5Yx65qD"
}
},
{
"cell_type": "markdown",
"source": [
"### Downloading the dataset from [Kaggle](https://www.kaggle.com/c/dogs-vs-cats) using API"
],
"metadata": {
"id": "k4_lNnfs696q"
}
},
{
"cell_type": "code",
"source": [
"!ls"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "QFSUjU6cd6dg",
"outputId": "ce6e53dc-dc41-44ab-80a4-b0db951c4af8"
},
"execution_count": null,
"outputs": [
{
"output_type": "stream",
"name": "stdout",
"text": [
"sample_data\n"
]
}
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "icKk9GKPZ-_w",
"outputId": "951dfa4e-4a9b-48d1-d717-8006ab325d1b"
},
"outputs": [
{
"output_type": "stream",
"name": "stdout",
"text": [
"Warning: Your Kaggle API key is readable by other users on this system! To fix this, you can run 'chmod 600 /root/.kaggle/kaggle.json'\n",
"Downloading dogs-vs-cats.zip to /content\n",
" 97% 790M/812M [00:04<00:00, 209MB/s]\n",
"100% 812M/812M [00:04<00:00, 187MB/s]\n"
]
}
],
"source": [
"!mkdir -p ~/.kaggle\n",
"!cp kaggle.json ~/.kaggle/\n",
"\n",
"!kaggle competitions download -c dogs-vs-cats"
]
},
{
"cell_type": "markdown",
"source": [
"### Extracting the ZipFile contents."
],
"metadata": {
"id": "HxGsPaoL7MXb"
}
},
{
"cell_type": "code",
"source": [
"import zipfile\n",
"zip_ref_1 = zipfile.ZipFile('/content/dogs-vs-cats.zip', 'r')\n",
"zip_ref_1.extractall('/content')\n",
"zip_ref_1.close()\n",
"\n",
"zip_ref_2 = zipfile.ZipFile('/content/train.zip', 'r')\n",
"zip_ref_2.extractall('/content')\n",
"zip_ref_2.close()"
],
"metadata": {
"id": "VlYByPLY6ba1"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "markdown",
"source": [
"### Importing the necessary libraries to work with in this notebook."
],
"metadata": {
"id": "PNESWB2p7Qyt"
}
},
{
"cell_type": "code",
"source": [
"# Importing libraries for numerical operations and data manipulation\n",
"import numpy as np\n",
"import pandas as pd\n",
"import random\n",
"random.seed(42) # Setting random seed for reproducibility\n",
"\n",
"# Importing libraries for file handling and system operations\n",
"import os, shutil, pathlib\n",
"\n",
"# Importing libraries for data visualization\n",
"import matplotlib.pyplot as plt\n",
"import plotly\n",
"import plotly.graph_objs as go\n",
"from PIL import Image\n",
"\n",
"# Importing libraries for deep learning\n",
"from tensorflow import keras\n",
"from tensorflow.keras import layers, models, callbacks\n",
"from tensorflow.keras.preprocessing.image import ImageDataGenerator, load_img\n",
"from tensorflow.keras.utils import image_dataset_from_directory\n",
"from tensorflow.keras.models import load_model\n",
"from tensorflow.python.keras.models import Sequential\n",
"\n",
"\n",
"# Importing libraries for machine learning model evaluation\n",
"from sklearn.metrics import confusion_matrix, classification_report, precision_recall_curve\n",
"\n",
"# Importing specific components from Keras\n",
"from keras.layers import Dense, Conv2D, Flatten, Dropout\n",
"# Setting up offline mode for Plotly\n",
"plotly.offline.init_notebook_mode()\n",
"\n",
"# Ignoring warnings\n",
"import warnings\n",
"warnings.filterwarnings(\"ignore\")"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/",
"height": 17
},
"id": "vXR8SSwufdXU",
"outputId": "77ab8d62-9219-4b21-875c-672e9fc79429"
},
"execution_count": null,
"outputs": [
{
"output_type": "display_data",
"data": {
"text/html": [
" \n",
" "
]
},
"metadata": {}
}
]
},
{
"cell_type": "markdown",
"source": [
"### Making Subsets - Train, Validation and Test"
],
"metadata": {
"id": "tzcg-RQt7XA3"
}
},
{
"cell_type": "code",
"source": [
"# Paths to the original dataset\n",
"original_dir = pathlib.Path(\"./train/train/\")\n",
"data_folder = pathlib.Path(\"./train/kaggle_dogs_vs_cats_small\")\n",
"\n",
"# Function to create a subset of the dataset with a specified range of files\n",
"def make_subset(subset_name, start_index, end_index):\n",
" for category in (\"cat\", \"dog\"):\n",
" dir = data_folder / subset_name / category # Path for the current category within the subset\n",
" os.makedirs(dir, exist_ok=True)\n",
"\n",
" fnames = [f\"{category}.{i}.jpg\" for i in range(start_index, end_index)]\n",
" for fname in fnames:\n",
" shutil.copyfile(src=original_dir / fname, dst=dir / fname)\n",
"\n",
"\n",
"# Total number of files\n",
"total_files = 4000\n",
"\n",
"# Calculate the number of files for validation and test subsets (15% each)\n",
"validation_size = int(0.15 * total_files)\n",
"test_size = int(0.15 * total_files)\n",
"\n",
"# Calculate the number of files for the training subset\n",
"train_size = total_files - validation_size - test_size\n",
"\n",
"# Create subsets for training, validation, and testing with specified start and end indices\n",
"make_subset(\"train\", start_index=0, end_index=train_size)\n",
"make_subset(\"validation\", start_index=train_size, end_index=train_size + validation_size)\n",
"make_subset(\"test\", start_index=train_size + validation_size, end_index=train_size + validation_size + test_size)\n"
],
"metadata": {
"id": "JER5sbRogeO7"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "markdown",
"source": [
"### Plotting sample images from the dataset randomly."
],
"metadata": {
"id": "sSba4otA7gM6"
}
},
{
"cell_type": "code",
"source": [
"num_samples = 9 # Images to plot\n",
"\n",
"# List all image filenames\n",
"image_filenames = os.listdir(original_dir)\n",
"\n",
"# Shuffle the list of image filenames randomly\n",
"random.shuffle(image_filenames)\n",
"\n",
"# Plotting the images\n",
"plt.figure(figsize=(12, 24))\n",
"for index in range(num_samples):\n",
" # Get the filename of the image at the current index\n",
" filename = image_filenames[index]\n",
"\n",
" # Determine whether the image is of a cat or a dog based on the filename\n",
" category = \"cat\" if \"cat\" in filename else \"dog\"\n",
"\n",
" # Load the image with a target size of (150, 150) (adjust target_size as needed)\n",
" img_path = os.path.join(original_dir, filename)\n",
" img = load_img(img_path, target_size=(150, 150))\n",
"\n",
" # Create a subplot and plot the image\n",
" plt.subplot(6, 3, index + 1)\n",
" plt.imshow(img)\n",
" plt.xlabel(filename + ' (' + category + ')' )\n",
"\n",
"plt.tight_layout()\n",
"plt.show()\n"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/",
"height": 1000
},
"id": "mp53g04YgY4e",
"outputId": "db7d9f3a-bba0-4f77-a7c3-933b21ef785a"
},
"execution_count": null,
"outputs": [
{
"output_type": "display_data",
"data": {
"text/plain": [
"| \n", " | filename | \n", "category | \n", "
|---|---|---|
| 0 | \n", "dog.1811.jpg | \n", "dog | \n", "
| 1 | \n", "cat.5411.jpg | \n", "cat | \n", "
| 2 | \n", "dog.12398.jpg | \n", "dog | \n", "
| 3 | \n", "cat.4161.jpg | \n", "cat | \n", "
| 4 | \n", "cat.6102.jpg | \n", "cat | \n", "
| \n", " | filename | \n", "label | \n", "
|---|---|---|
| 0 | \n", "cat | \n", "cat | \n", "
| 1 | \n", "dog | \n", "dog | \n", "